Production of Legal Rules

Edited by

Francesco Parisi

Oppenheimer Wolff and Donnelly Professor of Law, University of Minnesota, USA and Professor of Economics, University of Bologna, Italy

ENCYCLOPEDIA OF LAW AND ECONOMICS, SECOND EDITION

Edward Elgar Cheltenham, UK • Northampton, MA, USA

PART IV

INTERNATIONAL LAW

14 International law as a source of law Paul B. Stephan

1. Introduction The nature of international law as a legal system, and thus as a source of law, is controversial. At the conceptual level, the objection to characterizing international law as law is that it lacks a rule of recognition, i.e. there exists no agreement about the mechanism for determining which propositions of international law have the character of a legally binding obligation (Hart 1994: 213-37). At the practical level, the objection is the absence of a centralized enforcement authority (Scott and Stephan 2006: 5). Advocates of international law respond that substantial social conventions have emerged that constitute a rule of recognition, even if disagreement survives around the edges of the concept. They further assert that international law employs a range of enforcement mechanisms of varying degrees of centralization, including reputational effects, retaliation, and reciprocal rewards as well as some direct enforcement (Dunoff and Trachtman 1999; Guzman 2008; Scott and Stephan 2006; Sykes 2007; Trachtrnan 2008).

A functional approach to the problem of recognizing international law as a source of law can clarify both the analytic and evidential issues. Focusing on the characteristics and competence of the particular institution that might apply international law enables the identification of instances where international law operates as a significant constraint on state actors, although within limited domains. This approach requires distinguishing among several institutional settings. Treaty-based international institutions, whether an organ of the United Nations, the Dispute Settlement Body of the World Trade Organization (WTO), a tribunal constituted under a bilateral investment treaty and the Washington Convention on the Settlement of Investment Disputes between States and Nationals of Other States, or some other body, operate on the basis of particular treaties, general principles about the application and interpretation of treaties, and their own traditions as embraced or tolerated by affected parties. National tribunals will look to their particular constitutions, legal traditions, and their past practices as means of determining both what they will recognize as international law and when to apply those rules to disputes within their competence. National bureaucracies responsible for repeated interactions with their counterparts in other states (militaries constituting an important example, but also ministries of finance, foreign relations, environmental protection, and the like) will look to

255

256 Production of legal rules

their constituting authority (typically domestic statutes) and their past practices to determine what rules of international law they will recognize and apply. Finally, private actors, principally academics and advocates for particular interests, will construct and invoke international law to advance their causes. There is no reason to believe that the patterns of selection, interpretation and application of international law will converge across these functional groups, and much evidence to suggest that they diverge profoundly (Stephan 2010).

Scholars trained in economics find the functional approach sketched above congenial. It is only the last decade, however, that has seen much use of economic analysis to study the problem of legalization of international law. Political scientists to some extent have preceded economists in this endeavor. Their work combines rational-choice concepts drawn largely from economic theory with sociological insights that to some extent rest on assumptions and methodologies that oppose economic analysis (Keohane, Moravcsik and Slaughter 2001). Earlier studies of the problem, by contrast, rested entirely on a mix of sociology and moral theory (Chayes and Chayes 1995; Franck 1995).

2. Definition of International Law As a formal matter, international law derives from two processes, treaty-making and custom. Each of these mechanisms presents many difficult and controversial problems. With respect to treaties, there are persistent issues of interpretation and scope, as well as technical questions such as whether a treaty is in force and the means by which it can (as well as if it can) be suspended or terminated. Custom is said to result from state practice based on opinio juris, that is, a belief that the practice reflects a legal obligation rather than only a policy preference. There is no consensus, however, on what counts as state practice or how to determine the existence of opinio juris. International law specialists also debate whether there exist alternative sources of international law based on convergences in state practice that do not rise to the level of custom.

2.1 Treaties Treaties are express agreements among states that create legal obligations. The creation of treaties involves three steps: (1) making the treaty by representa­tives of the states that intend to become parties: (2) confirmation of the treaty by competent lawmakers in those states; and (3) ratification of the treaty by representatives of the state parties. The second step presents only questions of domestic law. In the United States, for example, the Constitution specifies that two-thirds of the Senate must consent to a treaty. 1 Whether a treaty has

1 The US judiciary, however, does not dictate which international agreements constitute treaties subject to this requirement. Scholars have analyzed the strategic opportunities this rule presents to the Executive (Hathaway 2008).

International law as a source of law 257

been made and ratified so as to give rise to legal obligations turns first on non-treaty international law, which is to say customary international law (CIL). As a secondary matter, however, a particular institutional actor (international tribunal, domestic court, domestic bureaucracy, or private actor) might look to domestic law as well to determine whether it will recognize a treaty as in force. Again looking to a US illustration, the Supreme Court has indicated that it will defer to the Executive in determining whether a treaty has entered into or remains in force.2 The Executive may derive the rules that it applies in such cases through the basis of bureaucratic interactions with other states, especially those among foreign ministries. But the US judiciary normally will not conduct an independent examination as to either the derivation or the application of those rules in cases where the Executive reaches a conclusion.

Application of treaties in force requires determinations of the meaning of the rules contained in a treaty (interpretation) and the extent of the treaty's domain (scope). International tribunals, domestic courts, domestic bureaucracies, and private actors have occasion to make such determinations, which each does based on its perception of the state of CIL, as well as its prior history in addressing such questions. There exist no formal mechanisms to induce the various actors that might apply treaties to conform to each other's determinations, and substantial evidence indicating that these determinations vary based on the actor involved as well as on context and interests (Stephan 2009). US courts assert that they give substantial deference to treaty interpretations proposed by the Executive, but they do not invariably follow them.3

States can affect ex ante determinations as to a treaty's interpretation and scope through formal reservations, understanding and declarations (RUDs) made at the time of a treaty's ratification. Some treaties provide expressly for such measures, but others do not. Two controversial issues follow: when are RUDs effective in the absence of treaty language providing for them? If a party has made an invalid R UD, is the legal consequence nonapplication of the treaty to that party, or instead application of the treaty in full without taking the invalid RUD into account?4

2 Charlton v Kelly, 229 US 447,474-5 (1913); Terlinden v Ames, 184 US 270,285 ( 1902). This issue is distinct from the question of whether the President may unilaterally terminate a treaty. See Goldwater v Carter, 444 US 996 (1979) (considering but not deciding the issue).

3 E.g., Chan v Korean Airlines, 490 US 122 (1989). 4 A related but distinct issue is whether a treaty maker has, as a matter of domestic

law, the capacity to make an RUD that will bind a domestic court. Compare Sosa v Alvarez-Machain, 542 US 692, 734 (2004) (treating reservation as legally binding on domestic court), with Igartua-de la Rosa v United States, 417 F. 3d 145, 186-9 (1st Cir. 2005) (Howard, J., dissenting) (arguing that US treaty makers lack constitutional authority to impose RUDs that would limit judicial power).

258 Production of legal rules

Some scholars have used economic arguments to support a liberal approach to RUDs. They argue that giving effect to RUDs in most cases where a treaty does not expressly prohibit them allows more states to take part in treaty regimes by overcoming barriers to ratification that would otherwise exist (Bradley and Goldsmith 2000). Others maintain both that RUDs create negative externalities by undermining regimes as a whole, and that refusing to give effect to an RUD while regarding a state as still bound by the unmodified treaty will not have a significant affect on treaty ratification (Goodman 2002). The latter argument rests fundamentally on claims about the socialization process and not on rational-actor models (Goodman and Jinks 2004). Yet another line of analysis argues that the threat of retaliatory and symmetrical RUDs deters states from invoking this mechanism as often as a simple rational-actor model might predict (Parisi 2009; Fon and Parisi 2008).

2.2 Custom Perhaps the greatest controversies in the field involve the practicality and the legitimacy of looking to custom as a source of binding international law. Traditional doctrine asserts that consistent state practice reflecting a sense of legal obligation can produce a binding norm of international law, even with respect to states that did not participate in the formation of the custom. The issue becomes especially fraught when a national legal system incorporates CIL into national law. In hierarchical national legal systems, such as that of the United States, there arise subsidiary but important questions such as (1) does CIL partake in the nature of federal law and thus pre-empt inconsistent state law? (2) does CIL limit the discretion of Congress to enact inconsistent statutes? (3) does CIL that arises after enactment of a congressional statute supersede that statute? and (4) do members of the Executive branch have the authority to act inconsistently with CIL (Bradley and Goldsmith 1997)?

Drawing on game theory, Goldsmith and Posner have attacked the premise that states frequently act out of a sense of legal obligation rather than out of self-interest. They argue that most instances where states exhibit behavioral regularities reflect common interest. They offer as an example the nineteenth­century custom of naval vessels during times of war not to interfere with peaceful fishing by their adversary's subjects. Interdiction of nonthreatening fishermen, they argued, diverts resources from combat for no appreciable gain. In instances where interests diverge, the possibility of working out a stable solution to a multiple-play, multiple-player game is small (Goldsmith and Posner 1999; 2005: 23-78).

Norman and Trachtman challenged Goldsmith and Posner's game-theory analysis. They argued that, under certain realistic assumptions, a stable solution to a multi-player, multiple-iteration prisoners' dilemma game could arise. This solution in turn would represent a custom based on obligation rather than

International law as a source of law 259

interest (Norman and Trachtman 2005; Trachtman 2008). Guzman offered a different challenge to Goldsmith and Posner. He maintained that states have an interest in a reputation for law-abiding behavior, and that this interest bridges the gap between rational-actor models of state behavior and the legal-obligation component of CIL (Guzman 2002, 2008). Scholars in turn criticized Guzman's account of state reputational interests. They observe that states have an interest in being seen as cooperative and other-regarding, but that it is incorrect to identify this interest with a reputation for law-abiding behavior. Much that international law permits is uncooperative, and much cooperative behavior is not required by international law (Downs and Jones 2002; Brewster 2009).

More generally, scholars have sought to ground theories of CIL in law­and-economics accounts of spontaneous law and informal norm enforcement (Parisi 2000; Fon and Parisi 2006, 2009; Swaine 2002; Verdier 2002). The parallels are obvious and the insights fruitful, but more work needs to be done on specifying (1) the mechanisms that lead states, as distinguished from individuals, to make particular choices, and (2) the relevant characteristics of the international community, as distinguished from communities populated by individuals (Dunoff and Trachtman 1999; Posner and Sykes 2007; Scott and Stephan 2006: 43-58). In particular, one can model state behavior as a two-level game in which domestic and state-to-state interactions proceed with each in mind (Putnam 1988). Individual behavior does not lend itself easily to this model. In other words, a successful theory of CIL needs a convincing theory of state decisionmaking with respect to international relations that takes into account the disaggregated interests of domestic actors. Accounts of custom within communities made up of individuals do not do this.

2.3 Other Sources There also is a debate over the existence of sources of international law other than treaties and custom. Many treaties constituting international tribunals, most prominently the Statute ofthe International Court of Justice, authorize the tribunal to invoke "the general principles oflaw recognized by civilized nations" as well as "judicial decisions and the teachings of the most highly qualified publicists of the various nations, as subsidiary means for the determination of rules of law."5 Prominent British treatises have argued from the fact that international tribunals can apply these rules that international law comprises these "general principles" as well as the "best practices" of national courts and the teachings of publicists (Brownlie 2003: 15-18; Jennings and Watts 1992: 36-40). Some international law professors in particular have found the reference to the "teachings of the most highly qualified publicists" as especially

5 Statute of the International Court of Justice, Art. 38(1)(c ), (d).

260 Production of legal rules

compelling and have sought to persuade other actors that their writings should be seen as a source of binding internationallaw.6

A more careful analysis, however, leads only to the conclusion that international tribunals, including ad hoc bodies set up to address particular disputes, generally have been provided with terms of reference that allow them to refer to those principles of national law that they find attractive. The two leading British treatises contain no examples of invocations of this principle that do not involve international tribunals. There is no basis to infer that this practice also authorizes national courts or national bureaucracies to find in general principles of national law a rule of decision that their own national law does not contain. Private actors, of course, are free to do what they wish but remain subject to reputational constraints.

The argument for allowing international tribunals to invoke sources other than treaties and CIL rests on the assertion that their mission is to resolve disputes, not to contribute to the development of a general body of law that binds other tribunals, much less other actors. There is no rule of recognition, comparable to the common law concept of precedent, that obliges a particular international tribunal to apply the same rules and to use the same reasoning as earlier tribunals, except in the relatively unusual case where the tribunal has a permanent charter and stable membership (such as the International Court of Justice). Under conditions of one-off dispute settlement, giving a tribunal a relatively free rein to choose from among potential legal rules complements its function to reach a reasonable settlement. The absence of precedential effect minimizes the risk that the decision made will generate negative effects on subsequent behavior (Benvenisti 2004). By contrast, scholars have argued that delegating comparable discretion to permanent international bodies presents serious agency problems (Stephan 2002).

One pathway to the "domestification" offoreign law, specifically the general principles recognized by "civilized nations," involves characterizing these principles as international law, and thus as a component of national law? Most

6 At least one US court has met this argument with indignation:

This notion - that professors of international law enjoy a special competence to prescribe the nature of customary international law wholly unmoored from legitimating territorial or national responsibilities, the interests and practices of States, or (in countries such as ours) the processes of democratic consent- may not be unique, but it is certainly without merit. (United States v Yousej, 327 F.3d 56, 102 (2d Cir. 2003))

7 An alternative mechanism has been to interpret particular provisions of domestic law as inviting incorporation of foreign standards. The US Supreme Court, for example, on occasion has looked to foreign practice as a guide to determine what constitutes "evolving standards of decency," which some members see as a component of the Eighth Amendment's prohibition of "cruel and unusual punishment." Other Justices criticize this practice. For the latest instance of this practice, see Graham v Florida, 130 S. Ct. 2011. (2010).

International law as a source of law 261

observers concede that the process of selecting principles of law found in domestic legal systems for elevation into international law involves subjective judgments and under-examined normative preferences. Some have argued that the practice of showing solidarity with other national legal systems strengthens the role of national courts, independent of the content of the rules over which they show solidarity. This safety-in-numbers argument in turn has been criticized on the grounds that judicial authority does not function as an end in itself and that some forms of judicial solidarity can be pernicious. For a review of the literature, see Stephan (2006).

In addition, the foreign solidarity argument is in tension with, if not directly opposed to, a conventional model found in the law-and-economics literature that portrays the judiciary as an agent of lawmakers (Posner 1982; Landes and Posner 197 5). The capacity to pick and choose among foreign laws and juristic opinions to find rules to apply amounts to an unlimited power to act as the rule-applier wishes, thereby breaking the bond between agent and principal. Granting domestic judges the authority to invoke foreign law when they find it more appealing than domestic law entails a significant departure from traditional explanations of the judicial function. Justification of this step rests more on normative moral arguments than on positive economic analysis.

3. Uses of International Law Recognition of a rule as belonging to the body of international law is only the first step in treating international law as a source of law. A competent body must then determine that it may apply that rule. This second step necessarily requires analysis of the applying body's authority.

Not all bodies have the authority to invoke all possible rules. Limitations such as standing requirements and choice of remedies effectively constrain which actors can invoke international law on whose behalf (Sykes 2005). Official institutions, whether international or domestic, have terms of reference that specify what they have the legal capacity to do. In some instances official authority is delegated to private actors, most frequently through the mechanism of arbitration. When exercising delegated authority, private actors work within the constraints of the delegation. In those instances where private actors function on their own account, they do not face formal constraints on their competence, but their determinations have consequences only to the extent that other actors choose to approve and follow their position. The pressure to win approval in turn acts as a constraint on what private actors can do.

3 .I International Institutions Many different kinds of international institutions have dispute-resolution authority that requires them to announce and apply rules of decision. For purposes of analysis, one can separate these institutions into five categories

262 Production of legal rules

- (1) nonjudicial institutions empowered to take actions that states by treaty are obligated to implement; (2) institutions that have the authority to identify lawbreakers and hence to "name and shame"; (3) institutions that may advance or withhold resources to states or private persons on the basis of legal determinations; (4) formal dispute resolution tribunals with settled jurisdiction and membership; and ( 5) ad hoc tribunals that acquire jurisdiction after a dispute has arisen. There are differences in the way institutions in each category invoke and apply international law.

3 .1.1 Nonjudicial bodies with authority to bind states The best example of a nonjudicial institution with authority to make decisions that by treaty bind states is the UN Security Council. Decisions of the Council, which typically deal with economic and military sanctions imposed on states that threaten international peace and security, are mandatory, in the sense that membership in the United Nations carries with it a treaty commitment to comply with those decisions.8

The European Commission (EC) and the European Parliament (EP) also fit into this category, as do comparable bodies in other regional organizations.

Because decisions of the Security Council are not subject to review by other organs of the United Nations and the five permanent members have a common interest in maintaining that body's authority, the Security Council has a tendency to make, rather than follow, intemationallaw.9 By contrast, institutions that face judicial review within their own treaty regime, as do the EC and the EP with respect to the European Court of Justice, face stronger limitations and tend to articulate more fully the legal basis for their actions.

3 .I .2 Nonjudicial bodies with authority to impose reputational sanctions A number of treaties creating substantive obligations in areas such as human rights and combatting corruption set up committees comprising representatives of the state parties. Typically these bodies review the performance of parties, and in some cases can entertain complaints by private persons. They issue reports, which can have reputational consequences (Helfer and Slaughter 2005: 924--8). There exist no other formal enforcement mechanisms that these bodies can

8 That a treaty obligation binds a state does not necessarily mean that, as a matter of domestic law, a national authority may implement a decision of the Security Council in advance of instructions from its parliament. See, e.g., Kadi v Council (Joined Cases C-402/05 and C-415/05), 2008 ECR I -6351 (refusing to enforce Security Council decision because of inconsistency with European law). Cf. Medllfn v Texas, 552 U.S. 491 (2008) (regarding capacity of Security Council to enforce decisions of International Court of Justice as undercutting argument that those decisions operate directly in US law in the absence of congressional authorization).

9 E.g., UN Sec. Counc. Res. 1757, S/RES/1757 (2007) (ordering treaty with Lebanon to enter into force in spite of Lebanon's failure to ratify).

International law as a source of law 263

employ, and correspondingly there are no formal means to obtain review of the their decisions. Casual empiricism suggests that some committee reports contain strong, if not unprecedented, claims about rules of international law, typically in a manner that promotes bootstrapping of jurisdiction.l0 The phenomenon has not been studied systematically, however.

Very little empirical work has been done on the effectiveness of name-and-shame international institutions. One study reported that a statistically significant negative correlation existed between a state's decision to join treaties that contain these mechanisms and the state's adherence to treaty standards, as measured by nongovernmental monitors (Hathaway 2002). This study provides some evidence to support the hypothesis that these councils are ineffective as a means of influencing state behavior. Critics of the study respond that it failed to follow state behavior along a sufficiently lengthy time horizon and therefore does not exclude the possibility that the combination of treaty adherence and council supervision produces cultural changes that eventually influence state behavior (Goodman and Jinks 2003).

3 .1.3 Nonjudicial bodies with authority to extend or withhold resources Some international institutions control valuable resources that states wish to obtain. Important examples are the various international financial institutions, including the International Bank for Reconstruction and Development (the World Bank) and the International Monetary Fund (IMF). These bodies make their decisions largely on the basis of business and policy criteria, but on occasion they also make determinations about the legality of a state's actions. Both the World Bank and the IMF, for example, will consider whether a state has complied with various international obligations, such as anti-corruption measures or environmental commitments, as a factor in their extension of financial resources (Stephan 1996-7). The Financial Action Task Force (FATF), a working group associated with the Organisation for Economic Co-operation and Development (OECD), determines whether states maintain sufficient safeguards against money laundering. A negative determination puts private banks on notice that a transfer of money to or from banks in the sanctioned country entails heightened risk of (national) prosecution for money laundering and thus induces a boycott of those banks.

A study of IMF regulation of national currency controls argues that that institution seldom exercises its authority to make legal determinations directly

IO E.g., General Comment No. 24, Issues Relating to Reservations made upon ratification of accession to the Covenant or the Optional Protocols thereto, or in relation to Declarations under Article 41 of the Covenant, CCPR/C/21/Rev.l/Add.6 (1994) (official document of Human Rights Committee maintaining that state consent was irrelevant to coverage of major human rights obligations).

264 Production of legal rules

(Simmons 2000). Instead, it relies on private capital markets to discipline state actors. IMF influence on state behavior operates indirectly through the tendency of private banks to respond to signals that it issues. A similar mechanism explains the influence of the FATF on state behavior. This limited evidence, when compared with the record of "name-and-shame" councils discussed above, is consistent with a conjecture that signals backed up by potential threats to withhold material resources induce a greater state response than do less tangible reputational threats.

A somewhat different but analogous instance of a resource-controlling international institution involves the Internet Corporation for Assigned Names and Numbers (ICANN). This entity, a nonprofit corporation established under US law with the blessing of the US government and, implicitly, the international community, controls the assignment of internet domain names. It exercises a dispute resolution function with respect to a species of trademark law, namely the ownership of domain names. Arbitration under ICANN auspices results in determinations that this body then applies through the effective sanction of refusing to recognize a disallowed name. Because ICANN controls the technical process that matches domain name with numeric addresses of internet sites, the sanction is immediate and absolute. Interested persons can relitigate their ownership claims in national courts, which have no obligation to respect the ICANN determinations. National litigation, however, involves considerably greater costs than does ICANN arbitration. Accordingly, ICANN exercises significant authority over economically significant ownership rights with only limited state supervision (Goldsmith and Wu (2006) at 168-71).

3 .1.4 Formal and ongoing international tribunals Formal international dispute-settlement tribunals have received considerable attention in recent years. These bodies are defined by their receipt of jurisdiction ex ante to consider particular categories of cases, as well as their stable structure and membership. Examples include the International Court of Justice, the Dispute Settlement Body of the WTO, the Tribunal for the Law of the Sea, and the International Criminal Court, as well as regional tribunals such as the European Court of Justice in Luxembourg (the court for the European Community) and the European Court of Human Rights in Strasbourg (the court for the Council of Europe, that is the parties to the European Convention on Human Rights).

The number of such tribunals has grown in the last two decades, leading some scholars to infer that they constitute successful substitutes for diplomatic resolution of international disputes (Helfer and Slaughter 2005). More skeptical observers identify potential mission creep as an agency cost that may lead states to decrease resort to some tribunals (Stephan 2002). The evidence is anecdotal rather than systematic, but does suggest that international tribunals that are perceived as idiosyncratic and self-aggrandizing lose business, largely

International law as a source of law 265

through states withdrawing acceptance of their jurisdiction (Posner 2009; Posner and Yoo 2005; Posner and Figueiredo 2005; Scott and Stephan 2006). The phenomenon is too recent, and the data set too small, to make conclusive assessments at this time.

Several scholars have documented the existence of autonomy among tribunals. Although a few tribunals have a two-stage structure analogous to the dyad of a first-instance court and an appellate court, no tribunals are constrained by any formal obligation to adopt the jurisprudence of any other. As a matter of practice, substantial divergences have arisen in the interpretation of the same substantive provisions of international law by different international tribunals (Brown 2007; Posner 2009; Stephan 2010). This practice provides weak evidence for the proposition that tribunals act in a way that reflects the particular contexts in which they operate and that a strong incentive to develop universal and general principles of international law across tribunals seems not to apply.

3 .1.5 Ad hoc international tribunals Ad hoc tribunals are those given terms of reference for settling a controversy that has already arisen. Both states and private actors invoke this mechanism. In some instances states and private actors make an ex ante commitment to use this mechanism but do not specify the members of the tribunal or its precise terms of reference until a specific controversy occurs. On other occasions, disputants will agree to use this mechanism in the absence of any prior commitment. Some formal and ongoing international tribunals, such as the International Court of Justice and the Permanent Court of International Arbitration, accept cases on an ad hoc basis. The Dispute Settlement Body of the WTO combines ad hoc tribunals, used for the first-instance consideration of a case, with a permanent tribunal that exercises appellate review.

Resort to ad hoc tribunals to address private international commercial disputes has grown significantly over the last two decades (Wood 2007). These bodies do not have to answer to any permanent international body for their actions. Constraints on their activity come in two forms. First, their awards are not self-executing, but rather depend on states to give them effect. Second, the disputants control selection of the members of the tribunal. There is substantial evidence that private parties take into account the records and reputations of potential arbiters. Arbiters in tum give some consideration to the fact that their awards will have value to parties only if states will give effect to them. For two particular categories of ad hoc dispute settlement -international commercial disputes and disputes between foreign investors and a host state - treaties oblige most states to honor these awards, but contain exceptions in instances of tribunal misconduct or overreaching, including the granting of awards that violate fundamental state interests. Arbiters thus

266 Production of legal rules

have an incentive to thread a line between pleasing the disputants and not antagonizing enforcing states by deviating too greatly from enforcing states' laws and policy (Posner 1999).

Ad hoc tribunals that hear disputes between states face more complex incentives. Selection to these tribunals involves complex networks of relationships, complicated by the possibility that government officials selecting tribunal members might someday become candidates for such appointments themselves. Unlike private arbitration, there is a greater possibility that the tribunal might award something more or other than a claim to money compensation. For example, the ad hoc tribunal operating under the auspices of the London Court of International Arbitration that resolved a long-term trade dispute between Canada and the United States required Canada to collect additional charges from lumber exporters .11 Actions of this sort require greater cooperation from the losing state than does an order to pay money, and to that extent constrains the arbiters' discretion to invoke new or surprising rules attributable to international law. Some research on ad hoc state-to-state tribunals indicates that this device enjoys growing popularity precisely because of such constraints (Posner and Yoo 2005).

3.2 National Courts National courts have treated international law as a source of law since time immemorial, but practice varies across states. The extent to which national courts invoke international law as well as the particular rules they choose depend on several variables, including relevant instructions from the respective national parliament to invoke international law, the particular policy area and its domestic context, the constitutional structure of the domestic court and its relationship with the domestic executive and parliamentary branches, and the international position of the state in question. The normative aspirations of the judges of these courts also have an obvious influence on these questions, although these do not readily lend themselves to economic analysis.

When judges act as agents of national lawmakers to implement instructions to apply international law, they must develop a methodology to determine the content of international law. In areas such as property boundary disputes and rules of contract, which often have the nature of a coordination problem rather than a prisoners' dilemma, a court might take a liberal view as to what constitutes applicable international rules. In other areas, however, the risk of foreign retaliation might be greater and a court correspondingly might demand clearer proof that an international rule exists before applying it pursuant to a

11 United States v Canada, LCIA, Case No. 7941, Award on Remedies, February 23,2009.

International law as a source of law 267

general legislative instruction. The behavior of the US Supreme Court is broadly consistent with this pattern.12

Early work explored the incentives that domestic judges face in deciding whether to derive rules of decision from international law. One model contrasted the possibility that judges might act as norm entrepreneurs with an alternative conception of judges as agents of a state engaged in playing a tit-for-tat game styled on the classic prisoners' dilemma (Stephan 1990). Judges who seek to function as norm entrepreneurs will invoke international law for one of two reasons. First, the judges may wish not simply to affect policy outcomes in their own jurisdiction, but to influence outcomes in other states. Second, the judges might not have available a domestic law that corresponds to their preferred outcome. In this latter case, the judges also need to construct a rule of hierarchy that allows a rule derived from international law to override domestic rules. Judges in federalist systems can elevate rules of international law by aligning them with national rules that trump subnational ones. Judges also can invoke the last-in-time convention to treat an international rule as overruling an earlier national rule.

The question of why judges might want to become international norm entrepreneurs is not well addressed by the economics literature. Sociological explanations focus on the existence of judicial networks, through which peers influence each other across borders (Slaughter 2004). Judges indeed might find offers of international travel and invitations to international meetings to be a valuable fringe benefit, given the strong restrictions most judges face on alternative sources of income. There is anecdotal evidence that national judges may take ambitious positions on the invocation of international law as a means of deflecting national limits on their behavior and expanding post-judicial career opportunities.l3 No systematic evidence exists indicating that such behavior occurs on a widespread basis, however.

Putting doctrinal and legitimacy concerns to the side, the principal economically based objection to judicial international norm entrepreneurship is its downstream effect on judicial practice. Lack of clarity about when judges will go outside domestic law and the mechanisms for locating international rules within a domestic legal hierarchy raise agency costs associated with

12 Compare Sosa v Alvarez-Machain, 542 US 692 (2004) (setting high standard for rules applicable under the so-called Alien Tort Statute), and Banco Nacional de Cuba v Sabbatino, 376 US 398 (1964) (refusing to recognize the CIL of international investor protection), with United States v California, 381 US 139 (1965) (offshore boundaries). For a survey of the post-war practice of the Court, see Stephan (2011).

13 A colorful example is Baltasar Garzon Real, who obtained a position with the International Criminal Court after Spain suspended him from his judicial position due to alleged misconduct based on his exertion of jurisdiction over controversial international law claims in apparent violation of national legislation.

268 Production of legal rules

the judiciary (Scott and Stephan 2006: 198-203). These costs in turn make international cooperation more costly by limiting the ability of representatives of national governments to specify the terms of their international commitments. Government representatives will find it more difficult to control the downstream consequences of an agreement if they cannot effectively constrain how their courts will implement it. Expansion of international norm entrepreneurial activity by domestic judges thus may impede state-to-state cooperation, including the development of international law based on international agreements.

Reviews of the application of international law by national courts indicate systematic variation in the rules invoked. These studies indicate that, rather than seeking to develop general and universal rules, national courts develop semi-autonomous international law regimes that respond to both the particular geopolitical context of the country in which they operate and their constitutional status within that country's legal regime (Bradford and Posner 2010). Benvenisti (2008) documents the way in which national courts invoke international law to protect their governments from international pressure. Stephan (2009) describes how the content of international law invoked at the national level corresponds to geopolitical power relations.

3.3 National Bureaucracies A great deal of state-to-state international interactions takes place through government bureaucracies with little or no supervision by domestic courts or international tribunals. Military bureaucracies, for example, develop handbooks on the laws of war, negotiate and implement arms control treaties, and impose discipline over soldiers, thereby developing and using international humanitarian law (also called the law of war, or jus in bello and jus ad bellum). Lawyers from foreign offices, competition authorities, financial regulators, environmental regulators, and other bureaucratic formations also meet regularly to work on various cooperative projects. Claims about international law suffuse this work, most of which are never tested outside of the bureaucracies themselves.

Over the last 15 years, a vast literature has emerged on international regulatory networks. It studies the mechanisms of cooperation among bureaucrats and the extent to which these interactions enable them to achieve independence from national constraints in the pursuit of multinational cooperative activity (Chayes and Chayes 1995; Slaughter 2004). This work largely focuses on processes of acculturation that stem from and further facilities cooperative conduct, including the development of rules that the actors regard as binding even when short -term national interests push in the opposite direction. Critics argue that these accounts are more aspirational than empirical and that the claim that international regulatory networks have a significant impact on the development of rules and policy remains unproved (Verdier 2009). Other research documents the extent to which international bureaucratic interactions correspond to conventional

International law as a source of law 269

rational-actor models without any need to impute a heightened preference for cooperation to the actors (Licht 1999).

Complex issues arise when advocates seek to persuade official actors to recognize international law developed by bureaucratic actors as providing a binding rule of decision. International tribunals, both permanent and ad hoc, work within particular terms of reference. The doctrinal device available to these bodies for refusing to apply a bureaucratically derived rule is to characterize it as lex specialis and therefore not available as general CIL. National courts also filter these arguments through their particular constitutional constraints and traditions.

3.4 Private Actors Private actors function as norm entrepreneurs regarding international law. Some serve as advocates for private disputants, some perform functions that complement (and perhaps even supersede) the "name-and-shame" function of international institutions, and some advance particular claims about international law through scholarship. Each of these activities has an impact on the development and application of international law, but only to the extent that these advocates can influence other actors exercising the authority to apply international law in a consequential manner.

Measuring the influence of private actors over persons possessing the authority to apply international law presents significant methodological challenges. No systematic study of the problem exists. One concept that has some salience is tournament theory, which studies competitive processes that result in the selection of winners and losers (Baker, Choi and Gulati 2006). Competition among private advocates for influence in the recognition of international law, as measured by adoption of positions advocated as well as the consequences of the adoption, is relatively easy to model and could be measured indirectly through observable data such as citations, positive outcomes in dispute resolution, and appointments to formal bodies. At present, however, no one has undertaken this task.

Bibliography Baker, Scott, Choi, Stephen J., and Gulati, Mitu (2006), "The Rat Race as an Information-Forcing

Device," 81 Indiana Law Journal53-82. Benvenisti, Eyal (2008), "Reclaiming Democracy: The Strategic Uses of Foreign and International

Law by Foreign Courts," 102 American Journal of International Law 241-74. Benvenisti, Eyal (2004), "Customary International Law as a Judicial Tool for Promoting Efficiency,"

in Benvenisti, Eyal and Hirsch, Moshe (eds.), The Impact of International Law on International Cooperation- Theoretical Perspectives, New York, NY: Cambridge University Press, 85-116.

Bradford, Anu and Posner, Eric A. (2011), "Universal Exceptionalism in International Law," 52 Harvard International Law Journal1-54.

Bradley, Curtis A. and Goldsmith, Jack L. (2000), "Treaties, Human Rights, and Conditional Consent," 149 University of Pennsylvania Law Review 399-468.

270 Production of legal rules

Bradley, Curtis A. and Goldsmith, Jack L. (1997), "Customary International Law as Federal Common Law: A Critique of the Modern Position," 110 Harvard Law Review 815-76.

Brewster, Rachel (2009), "Unpacking the State's Reputation," SO Harvard Journal of International Law 271-321.

Brown, Chester (2007), A Common Law of International Adjudication, New York, NY: Oxford University Press.

Brownlie, Ian (2003), Principles of Public International Law, 6th ed., New York, NY: Oxford University Press.

Chayes, Abram and Chayes, Antonia Handler (1995), The New Sovereignty: Compliance with International Regulatory Agreements, Cambridge, MA: Harvard University Press.

Downs, George W. and Jones, Michael A. (2002), "Reputation, Compliance, and International Law," 31 Journal of Legal Studies S95-ll4.

Dunoff, Jeffrey L. and Trachtman, Joel P. (1999), "Economic Analysis of International Law," 24 Yale Journal of International Law 1-59.

Fon, Viney and Parisi, Francesco (2009), "Stability and Change in International Customary Law," 17 Supreme Court Economic Review 279-309.

Fon, Viney and Parisi, Francesco (2008), "The Hidden Bias of the Vienna Convention on the Law of Treaties," 4 Review of Law and Economics 383-406.

Fon, Viney and Parisi, Francesco (2006), "International Customary Law and Articulation Theories: An Economic Analysis", 2 Brigham Young University International Law & Management Review 201-32.

Franck, Thomas M. ( 1995), Fairness in International Law and Institutions, New York, NY: Oxford University Press.

Goldsmith, Jack L. and Posner, Eric A. (2005), The Limits of International Law, New York, NY: Oxford University Press.

Goldsmith, Jack L. and Posner, Eric A. (1999), "A Theory of Customary International Law," 66 University o.fChicago Law Review 1113-77.

Goldsmith, Jack L. and Wu, Tim (2006), Who Controls the Internet? Illusions of a Border less World, New York, NY: Oxford University Press.

Goodman, Ryan (2002), "Human Rights Treaties, Invalid Reservations, and State Consent", 96 American Journal of International Law 531-60.

Goodman, Ryan and Jinks, Derek (2004), "How to Influence States: Socialization and International Human Rights Law," 54 Duke Law Journal62l-703.

Goodman, Ryan and Jinks, Derek (2003), "Measuring the Effects of Human Rights Treaties," 14 European Journal of International Law 171-200.

Guzman, Andrew T. (2008), How International Law Works- A Rational Choice Theory, New York, NY: Oxford University Press.

Guzman, Andrew T. (2002), "A Compliance-Based Theory of International Law," 90 California Law Review 1823-87.

Hart, H.L.A. (1994), The Concepto,fLaw -Second Edition, New York, NY: Oxford University Press. Hathaway, OonaA. (2008), "Treaties' End: The Past, Present, and Future oflnternational Lawmaking

in the United States," 117 Yale Law Journal1236-373. Hathaway, OonaA. (2002), "Do Human Rights Treaties Make a Difference?" 111 Yale Law Journal

1935-2042. Helfer, Laurence R. and Slaughter, Anne-Marie (2005), "Why States Create International Tribunals:

A Response to Professors Posner and Yoo," 93 California Law Review 901-56. Jennings, Robert and Watts,Arthur (1992), 1 Oppenheim's International Law, 9th edition, Harlow,

England: Longman Group UK Limited. Keohane, Robert 0., Moravcsik, Andrew, and Slaughter, Anne-Marie (2001), "Legalized Dispute

Resolution: Interstate and Transnational," in Goldstein, Judith L., Kahler, Miles, Keohane, Robert 0. and Slaughter Anne-Marie (eds.), Legalization and World Politics, Cambridge, MA: MIT Press, 73-104.

Landes, William M. and Posner, Richard A. (1975), "The Independent Judiciary in an Interest-Group Perspective," 18 Journal of Law & Economics 875-901.

Licht, Amir N. (1999), "Games Commissions Play: 2x2 Games of International Securities Regulation," 24 Yale Journal of International Law 61-128.

International law as a source of law 271

Norman, George and Trachtman, Joel (2005), "The Customary International Law Game," 99 American Journal of International Law 541-80.

Parisi, Francesco (2009), "The Economics of Treaty Ratification," 5 Journal of Law Economics and Policy 209-24.

Parisi, Francesco (2000), "Spontaneous Emergence of Law: Customary Law," in Bouckaert, Boudewijn and De Geest, Gerrit (eds.), 5 Encyclopedia of Law and Economics, Cheltenham, UK, Edward Elgar, 603-30.

Posner, Eric A. (2009), The Perils of Global Legalism, Chicago, IL: Chicago University Press. Posner, Eric A. ( 1999), "Arbitration and the Harmonization of International Commercial Law -A

Defense of Mitsubishi," 39 Virginia Journal of International Law 647-70. Posner, Eric A. and Figueiredo, Miguel F.P. (2005), "Is the International Court of Justice Biased?"

34 Journal of Legal Studies 599-630. Posner, Eric A. and Sykes, Alan 0. (2007), "An Economic Analysis of State and Individual

Responsibility under International Law", 9 American Law & Economics Review 72-129. Posner, Eric A. and Yoo, John C. (2005), "A Theory of International Adjudication," 93 California

Law Review 1-74. Posner, Richard A. ( 1982), "Economics, Politics, and the Reading of Statutes and the Constitution,"

49 University of Chicago Law Review 263-91. Putnam, Robert D. (1988), "Diplomacy and Domestic Politics: The Logic of Two-Level Games,"

42 International Organization 427-60. Scott, Robert E, and Stephan, Paul B. (2006), The Limits of Leviathan- Contract Theory and the

Enforcement of International Law, New York, NY: Cambridge University Press. Simmons, Beth A. (2000), "Money and the Law: Why Comply with the Public International Law

of Money?" 25 Yale Journal of International Law 323-62. Slaughter, Anne-Marie (2004), A New World Order, Princeton, NJ: Princeton University Press. Stephan, Paul B. (2010), "Disaggregating Customary International Law," 21 Duke Journal of

Comparative & International Law 191-205. Stephan, Paul B. (20 11), "Treaties in the Supreme Court: 1946-2000 ," in Sloss, David L., Ramsey,

Michael D. and Dodge, William S. (eds.), International Law in the U.S. Supreme Court: Continuity and Change, New York, NY: Cambridge University Press, 317-52.

Stephan, Paul B. (2009), "Symmetry and Selectivity: What Happens in International Law When the World Changes," 10 Chicago Journal of International Law 91-123.

Stephan, Paul B. (2006), "Process Values, International Law, and Justice," 23 Social Philosophy and Policy 131-52.

Stephan, Paul B. (2002), "Courts, Tribunals and Legal Unification- The Agency Problem," 3 Chicago Journal of International Law 333-52.

Stephan, Paul B. (1996-97), "Accountability and International Lawmaking: Rules, Rents and Legitimacy," 17 Northwestern Journal of International Law and Business 681-735.

Stephan, Paul B. (1990), "International Law in the Supreme Court," 1990 Supreme Court Review 133-61.

Swaine, Edward T. (2002), "Rational Custom," 52 Duke Law Journal559-627. Sykes, Alan 0. (2007), "International Law," in Polinsky,A. Mitchell and Shavell, Steven (eds.), 1

Handbook of Law and Economics, Amsterdam, Netherlands: Elsevier, 757-826. Sykes, Alan 0. (2005), "Public versus Private Enforcement of International Economic Law:

Standing and Remedy," 34 Journal of Legal Studies 631-66. Trachtman, Joel P. (2008), The Economic Structure of International Law, Cambridge, MA: Harvard

University Press. Verdier, Pierre-Hughes (2009), "Transnational Regulatory Networks and Their Limits," 34 Yale

Journal of International Law 113-172. Verdier, Pierre-Hughes (2002), "Cooperative States: International Relations, State Responsibility

and the Problem of Custom," 42 Virginia Journal of International Law 839-67. Wood, Diane P. (2007), "Private Dispute Resolution in International Economic Law," in Guzman,

Andrew T. and Sykes, Alan 0. (eds.), Research Handbook in International Economic Law, Cheltenham, UK, and Northampton, MA, USA: Edward Elgar, 575-97.